Developing Inhibitors of the p47phox-p22phox Protein-Protein Interaction by Fragment-Based Drug Discovery

Article abstract of DOI:10.1021/acs.jmedchem.9b01492

Nicotinamide adenine dinucleotide phosphate oxidase isoform 2 is an enzyme complex, which generates reactive oxygen species and contributes to oxidative stress. The p47phox-p22phox interaction is critical for the activation of the catalytical NOX2 domain, and p47phox is a potential target for therapeutic intervention. By screening 2500 fragments using fluorescence polarization and a thermal shift assay and validation by surface plasmon resonance, we found eight hits toward the tandem SH3 domain of p47phox (p47phox^SH3A-B) with K_D values of 400-600 μM. Structural studies revealed that fragments 1 and 2 bound two separate binding sites in the elongated conformation of p47phox^SH3A-B and these competed with p22phox for binding to p47phox^SH3A-B. Chemical optimization led to a dimeric compound with the ability to potently inhibit the p47phox^SH3A-B-p22phox interaction (K_i of 20 μM). Thereby, we reveal a new way of targeting p47phox and present the first report of drug-like molecules with the ability to bind p47phox and inhibit its interaction with p22phox.

Full text of DOI:10.1021/acs.jmedchem.9b01492

Subscriber access provided by UNIV OF TEXAS DALLAS
Article
Developing Inhibitors of the p47phox-p22phox Protein-
Protein Interaction by Fragment-Based Drug Discovery
Sara Marie Øie Solbak, Jie Zang, Dilip Narayanan, Lars Jakobsen Høj,
Saskia Bucciarelli, Charlotte Softley, Sebastian Meier, Annette Eva
Langkilde, Charlotte Held Gotfredsen, Michael Sattler, and Anders Bach
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 10 Jan 2020
Downloaded from pubs.acs.org on January 10, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Developing Inhibitors of the p47phox-p22phox Protein-Protein
Interaction by Fragment-Based Drug Discovery
†
†
†
†
†
Sara Marie Øie Solbak, Jie Zang, Dilip Narayanan, Lars Jakobsen Høj, Saskia Bucciarelli, Charlotte
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
#,‡
ǁ
†
ǁ
#,‡
Softley, Sebastian Meier, Annette Eva Langkilde, Charlotte Held Gotfredsen, Michael Sattler and
†,
Anders Bach *
†Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of
Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark. E-mail: anders.bach@sund.ku.dk
#Institute of Structural Biology, Helmholtz Zentrum München, 85764 Neuherberg, Germany
‡Biomolecular NMR and Center for Integrated Protein Science Munich at Department of Chemistry,
Technical University of Munich, 85747 Garching, Germany.
ǁDepartment of Chemistry, Technical University of Denmark, Kemitorvet, 2800 Kgs Lyngby, Denmark
ABSTRACT
NADPH oxidase isoform 2 is an enzyme complex, which generates reactive oxygen species (ROS) and
contributes to oxidative stress. The p47phox-p22phox interaction is critical for activation of the catalytical
NOX2 domain, and p47phox is a potential target for therapeutic intervention. By screening 2,500 fragments
using fluorescence polarization (FP) and a thermal shift assay (TSA) and validation by surface plasmon
resonance (SPR) we found eight hits towards the tandem SH3 domain of p47phox (p47phoxSH3A-B) with KD
values of 400–600 μM. Structural studies revealed that fragments 1 and 2 bound two separate binding sites
in the elongated conformation of p47phoxSH3A-B and these competed with p22phox for binding to
p47phoxSH3A-B. Chemical optimization led to a dimeric compound with the ability to potently inhibit the
p47phoxSH3A-B-p22phox interaction (K of 20 µM). Thereby, we reveal a new way of targeting p47phox and
i
present the first report of druglike molecules able to bind p47phox and inhibit its interaction with p22phox.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
The multisubunit enzyme complex NADPH oxidase isoform 2 is highly expressed in phagocytes where it
•-
catalyzes the reduction of molecular oxygen to superoxide anion (O ). This reactive oxygen species (ROS)
2
generating enzyme is known to have a key role in innate immunity, as demonstrated in chronic
granulomatous disease (CGD) caused by an inherited deficiency in one of the NADPH oxidase 2 subunits,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
where the phagocytes fail to kill invading microorganisms. Excessive ROS production by NADPH oxidase 2
on the other hand contributes to oxidative stress and is connected to a wide range of diseases involving
inflammation, diabetes, and cancer.2, 3 Furthermore, NADPH oxidases contribute to brain damage in
relation to traumatic brain injury, reperfusion injury in stroke, and neurodegenerative diseases.1, 3-5
The NADPH oxidase 2 complex is comprised of a membrane-bound flavocytochrome b558, a
heterodimer of the catalytically active NOX2 domain and p22phox, three cytosolic proteins, p40phox,
p47phox and p67phox, and a small G protein Rac GTPase.3, 6 NOX2 is activated by Ca -influx activating the
phosphokinase C (PKC), which further phosphorylates C-terminal serine residues located in the
autoinhibitory region (AIR) and the proline-rich region of p47phox.4, 7 The phosphorylation causes a
conformational change of p47phox, allowing the cytosolic proteins to relocate to the plasma membrane
where they associate with flavocytochrome b558, forming the active NADPH oxidase 2 complex. In its
inactive state, the p47phox tandem Src Homology 3 (SH3) domain (SH3A-B) is occupied by the AIR of
p47phox. However, upon p47phox phosphorylation in the AIR domain, the conformational change involves
detachment of the AIR domain from the SH3A-B domain, allowing SH3A-B to bind to the proline-rich
domain (PRD) of p22phox instead (Figure 1). The truncated p47phox SH3A-B (p47phox151–286) without the
AIR (here referred to as p47phoxSH3A-B) is considered a model protein of activated p47phox, as it has an
open tandem SH3 domain and does not require phosphorylation to bind to p22phox.8, 9 The structure of the
2+
8
tandem SH3 domain including the AIR (p47phox151-342; here referred to as p47phox ) and p47phox
AIR
SH3A-B
in
complex with the p22phox derived peptide p22phox149-168 have previously been investigated by X-ray
crystallography, NMR, and small angle X-ray scattering (SAXS).8-12 By X-ray crystallography, both p47phoxAIR
and p47phoxSH3A-B in complex with p22phox149-168 were found to form intertwined dimers.8, 12 However, in
2
ACS Paragon Plus Environment

Page 3 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
solution both proteins were found by NMR and SAXS analysis to be monomeric.9, 10 Interestingly, the SAXS
studies revealed an overall more compact conformation of the p47phoxSH3A-B-p22phox complex than for the
extended apo p47phoxSH3A-B. Furthermore, the binding interface of p47phoxSH3A-B to p22phox149-168 was
1
15
characterized by NMR chemical shift perturbation (CSP) analysis based on H- N HSQC NMR titration
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
93
194
204
experiments. In particular, for the three tryptophan residues in SH3A (Trp , Trp and Trp ) and the two
in SH3B (Trp263 and Trp ), significant CSPs were observed during titration with p22phox149-168. This finding
reflects that p22phox149-168 stably interacts with a region composed by both SH3 domains forming a crevice
in the middle of the tandem SH3 domain, as shown by NMR structure determination (PDB ID: 1WLP).9
264
Figure 1. p47phox-p22phox binding model. PKC activates p47phoxAIR by phosphorylating seven C-terminal
serine residues (three in the AIR domain), leading to a conformational change where the AIR domain is
removed from the tandem SH3 domain allowing binding to the PRD of p22phox.
The binding between p47phox and p22phox is essential for NOX2 activation, and p47phox has thus been
suggested as a target for therapeutic intervention.4, 13 This approach has the potential to provide novel,
specific, and genuine NOX2 inhibitors, which have been shown notoriously difficult to develop by other
1
4
strategies. However, the only small molecular compounds reported to prevent NADPH oxidase 2 activity
by direct inhibition of the p47phox-p22phox interaction are ebselen (2-phenyl-1,2-benzisoselenazol-3(2H)-
1
3
one) and a series of close analogs. Ebselen is a synthetic seleno-organic ROS scavenging compound known
to induce cysteine-targeted oxidation followed by S-thiolation - an oxidative reversible modification of the
cysteine residues of proteins.15
In this study, we show that ebselen binds p47phoxSH3A-B covalently, leading to destabilization and
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
aggregation of the protein – a mechanism that is not promising for further drug development. Instead, to
find novel reversible inhibitors of the p47phox-p22phox interaction, we used a fragment-based drug
discovery (FBDD) approach. We thereby identified several small-molecules binding to the p47phoxSH3A-B, the
best ones being dimeric compounds which display a 13–30-fold higher affinity relative to the parent
fragment hits.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Assay development and mechanism of ebselen.
In order to discover small molecular inhibitors of the p47phox-p22phox interaction, we developed an FP
inhibition assay, a TSA and an SPR binding assay, which were validated by testing the p22phox derived
peptides p22phox151-162, p22phox149-168, and p22phox149-162, all containing the PRD known to bind
p47phoxSH3A-B. The binding affinity of the p22phox peptides to p47phoxSH3A-B determined from FP and SPR
corresponded to literature values,8, 9, 13 and TSA showed increased melting curves indicative of binding
(Supporting Information Figures S1-3 and Table S1). According to literatures, SPR binding experiments of
p22phox peptides to p47phoxAIR gave a clear reduced binding affinity (Supporting Information Figures S3D-
E and Table S1).8, 16
Ebselen was found to prevent NOX2 activation by inhibiting the interaction between p47phox and
1
3
p22phox, as shown by FP. We therefore considered if ebselen could be a useful reference compound for
our assays and investigated the mechanism of action in more detail. In FP, ebselen in the absence of the
reducing agent tris-(2-carboxyethyl)phosphine (TCEP) yielded a K of 63 ± 6 μM, while a >10-fold increase in
i
Ki (>300 μM) was observed with TCEP present in the buffer, which indicate a covalent binding mode. When
prolonging the incubation time from 10 minutes to 2 hours in the FP assay without TCEP, the K was
i
reduced to 12 ± 0.07 μM (Figure 2A). In SPR, ebselen strongly bound p47phoxSH3A-B with an almost non-
reversible binding kinetic (slow dissociation rate) (Figure 2B). The binding was however reversed by TCEP
(
Figure 2B and Supporting Information Figure S4A). Also, LC-MS analysis of p47phoxSH3A-B with ebselen
showed a total mass of 17573.1 Da (Figure 2C), which compared to the mass of p47phoxSH3A-B without
ebselen (17299.5 Da) (Supporting Information Figure S4B) corresponds to a mass-increase similar to the
4
ACS Paragon Plus Environment

Page 5 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
molecular weight of ebselen (274.2 Da). These results demonstrate a covalent interaction of ebselen with
196
SH3A-B
the only cysteine (Cys ) of p47phox
(Figure 2D-E). Furthermore, SAXS studies indicated aggregation of
p47phoxSH3A-B in the presence of ebselen, evident by a strong increase in the maximum dimension of the
particles (Dmax) (starting at 235 Å, compared to the 86 Å without ebselen). The corresponding SAXS-derived
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
radius of gyration (R =67.6 Å) and the concentration-normalized scattering intensity at q=0 (I(0)/c =17.2
g
a.u) were also higher than without ebselen (26.5 Å and 1.1 a.u, respectively) (Figure 2F-G and Supporting
Information Table S2). The aggregation was observed shortly after sample preparation and increased over
time (data not shown).
Thus, ebselen inhibits NOX2 by covalent binding to Cys196 of p47phoxSH3A-B, leading to destabilization and
aggregation of p47phoxSH3A-B, rather than by occupying the p22phox binding pocket. This dramatic effect
can be explained by Cys196 being localized directly adjacent to residues Trp , Trp , Phe and Val186,
193
204
195
9
reported to form the hydrophobic binding pocket for p22phox in the p47phox SH3A domain. Although this
mechanism of action of ebselen prevents p47phox from binding to p22phox, resulting in reduced NOX2
1
3
17-19
activity as described, and ebselen may have some promise as a drug candidate,
such a mechanism
20
does not seem useful for a chemical probe, as it likely comes with unspecific effects. Indeed, ebselen has
been found active against a wide range of proteins, and is thus a non-selective compound.15, 21 Based on
this, we used the p22phox151-162 and p22phox149-168 peptides as assay controls.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Interaction of p47phoxSH3A-B with ebselen. A) Ebselen tested in the FP competition assay using the
Cy5-p22phox149-162 probe in the absence of TCEP – after 10 minutes (green) or 2 hours (blue) incubation –
and in the presence of TCEP (red). B) SPR sensorgram of ebselen (12.5 µM) injected over 4684 RU
immobilized p47phoxSH3A-B on a biosensor chip followed by abrogation of the complex by 2 mM TCEP. C) LC-
MS data of 0.2 mg/mL p47phoxSH3A-B with 200 µM ebselen showing a total mass of 17573.1 ± 0.5 Da. D) The
solution NMR structure of p47phoxSH3A-B in complex with p22phox149-168 (PDB ID: 1WLP); Cys196 in the SH3A
domain is highlighted (purple), while the p22phox peptide is removed. E) Our data suggest that ebselen
forms a selenylsulphide bond with Cys196 of p47phoxSH3A-B forming a stable complex. F) SAXS data for
p47phoxSH3A-B (black) and p47phoxSH3A-B in the presence of ebselen (red) plotted as concentration-
6
ACS Paragon Plus Environment

Page 7 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
normalized scattering intensity I(q)/c in arbitrary units vs. q and G) the corresponding pair-distance
distribution function P(r).
Fragment screening and hit validation.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
To develop direct inhibitors of the p47phox-p22phox interaction, 2,500 rule of 3 compliant fragments22
were screened at 1 mM for binding to p47phoxSH3A-B in the FP assay using the p22phox-derived Cy5 probe
(
Cy5-p22phox149-162), and at 0.5 mM for inducing thermal stabilization of p47phoxSH3A-B by TSA. The FP
screening resulted in 75 fragments (3.0% hit rate) that caused at least a 20% decrease in polarization
relative to the DMSO control. The TSA screen resulted in 109 fragment hits (4.4% hit rate), where a hit was
defined as a fragment that induced a ∆T > 0.9°C relative to the negative control, and a sigmoidal melting
m
curve with a fluorescence window in the proximity of the controls.
The 75 hits from the primary FP screen were validated in a 6-point FP dose-response experiment (63 µM
– 2 mM) in the absence and presence of 0.01% Triton-X to exclude aggregation-based promiscuous
inhibitors.23 The fragments were also tested using the TAMRA-probe (TAMRA-p22phox151-162), in the
absence of p47phoxSH3A-B, and by assessing the stability of the total fluorescence intensity (FLINT), in order
2
4
to remove false-positives due to fluorescence inner-filter effects. A total of 22 fragments had positive
results in at least three out of four tests and were selected as validated FP hits (0.9% hit rate) (the FP
validation data for 1 are shown in Figure 3A). The hits from the primary TSA screen were validated in a 4-
point TSA dose-response experiment (0.25-2 mM). A total of 98 fragments (3.9% hit rate) showed a dose-
dependent increment in ∆T , and were thus considered validated TSA hits (the TSA data for 1 are shown in
m
Figure 3B). Only three hits were validated in both FP and TSA. A low number of common hits is often seen
in fragment screening studies and is likely due to the underlying assay principles and sensitivities.25-27
Twenty-four of the validated fragments contained potentially reactive groups, such as thiourea and
nitriles, or had structural similarity to ebselen. In analogy to ebselen, one fragment, (2-(3-
(trifluoromethyl)phenyl)isothiazol-3(2H)-one), showed a mass increase in LC-MS equal the molecular
weight of the fragment (245.2 Da), indicating covalent binding, so this was removed from further studies
leaving 116 fragment hits validated in FP or TSA (Figure 3C).
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Fragment screening and validation. A) FP assay for fragment hit 1 competing with the fluorescent
probe Cy5-p22phox149-162 (red) and TAMRA-p22phox151-162 (purple) for binding to p47phoxSH3A-B. The assay
was also performed in presence of 0.01% Triton-X (green) and without protein (black). B) TSA melting
curves for p47phoxSH3A-B with various concentrations of 1. C) Venn diagram of validated screen hits from FP
and TSA. D-F) Validated FP and TSA hits were injected over immobilized p47phoxSH3A-B and evaluated by
SPR. Fragments showing a concentration dependent binding response and a fragment-like binding kinetic
(as 1, D) were confirmed as hits. Fragments showing no binding up to 1 mM (as example in E) or a non-
fragment-like binding behavior (as in F) were not regarded as true hits. G, H) Hits confirmed by SPR were
further investigated in an SPR inhibition assay by injecting 1 mM fragment over immobilized p47phoxSH3A-B
(
3358 RU) in the absence G) and presence H) of 10 µM p22phox149-168 in the running buffer. I) Venn diagram
8
ACS Paragon Plus Environment

Page 9 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of final validated screen hits from FP, TSA and SPR J) Chemical structures of the eight fragments selected
for further investigations.
Fragment binding to p47phoxSH3A-B characterized by SPR.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The 116 validated hits were further assayed for direct binding to p47phoxSH3A-B by SPR, where the fragments
were injected in two-fold serial dilutions over p47phoxSH3A-B immobilized on the biosensor chip. Seventy-
two fragments showed a concentration dependent response, thus confirming interaction with p47phoxSH3A-
B
(SPR data of 1 are shown in Figure 3D), while 39 fragments displayed low or no binding and five fragments
showed non-fragment like binding with slow on- or off-rates (see examples in Figure 3E-F).
To evaluate whether the 72 SPR validated fragments bind to or interfere with the binding interface of
2
8
p22phox at p47phox, we performed an SPR inhibition assay. The fragments (1 mM) were injected over
immobilized p47phoxSH3A-B, where the binding interface between p47phoxSH3A-B and p22phox was blocked
by including p22phox149-168 at saturating concentration (10 µM) in the running and ligand buffer. Following
this assay, a similar injection was performed over the p47phoxSH3A-B surface not blocked by p22phox149-168. A
reduced binding of the fragments to p47phoxSH3A-B blocked by p22phox, would indicate that the fragments
bind to the same interface at p47phoxSH3A-B as p22phox. Out of the 72 tested fragments, 14 showed a
reduced binding response in the inhibition assay, with a calculated metric S ≥0.15, corresponding to a
reduction in response level of ~25% (data for p22phox151-162 and fragment 1 are shown in Supporting
Information Figure S5 and Figure 3G-H, respectively). The 14 fragment hits were checked for structural
similarities to known pan-assay interference compounds (PAINS) or aggregators using the filtering tool
FAFdrugs4 (Free ADME-Tox Filtering Tool)29 and Aggregator Advisor.30 One fragment containing a
thioketone was flagged as a potential covalent binder. The SPR binding experiments and inhibition assays
of the remaining 13 hits were repeated in order to confirm these. Two of the fragments showed only little
reduction in binding in the presence of blocking peptide and were excluded, resulting in 11 SPR validated
hits (cf. Venn diagram in Figure 3I). Out of these, three had very weak affinities to p47phox, and thus eight
fragments (Figure 3J) were selected for further investigations by NMR and SAXS. SPR sensorgrams and
biophysical data for these are shown in Supporting Information Figure S6 and Table 1.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1. Biophysical data of the final 8 validated fragments. ∆T was calculated from TSA data, K from FP,
m
i
steady state affinities (K ) from SPR sensorgrams seen in Supporting Information Figure S6.
D
∆
T
m
(°C)
Compds Compd identity
1
MW
K
i
(µM)a
K
D
(µM)a
Sb
(
0.25/0.5/1/2 mM)
144.2 0.3/0.6/1.2/2.0
227.3 0.7/1.4/2.1/3.6
428 ± 23
606 ± 25
488  130
412  130
0.61  0.2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0.48  0.1
3
4
5
204.3 0.4/0.4/1.7/3.0
206.3 1.0/0.9/1.4/2.2
128.2 0.8/0.8/0.9/2.4
No hit
No hit
No hit
465  160
563  320
349  29
0.25  0.07
0.38  0.2
0.75  0.3
6
7
8
233.4 0.6/0.8/1.6/2.9
233.4 0.5/0.3/1.1/2.0
No hit
No hit
No hit
548  160
418  83
504  110
0.22  0.05
0.36  0.06
0.35  0.10
219.3 -0.1/0.3/0.9/1.4
aK
and K
values are reported as the mean ± SEM (n ≥ 3); S values are reported as mean ± SEM (n = 2)
b
i
D
NMR chemical shift changes in p47phoxSH3A-B upon fragment binding.
The SPR inhibition assay shows that p22phox reduces binding of the fragments, which indicates that our
fragment hits bind to the same site of p47phoxSH3A-B as p22phox. To explore this further, we analyzed the
1
15
CSPs in H- N HSQC spectra focusing on the five Trp key residues considered in the binding interface of
p47phoxSH3A-B to p22phox (Figure 4A). As described in the literature, large chemical shift changes with slow
exchange at the NMR chemical shift time scale were observed for all five well-resolved Trp indole NHs of
p47phoxSH3A-B when titrated with p22phox149-168 (p47phoxSH3A-B:p22phox149-168 ratios 1:0, 2:1, 1:1, 1:2 and
9
1
:4) (Figure 4B-C and Supporting Information Figure S7). The assignment of the Trp indole signals of
p47phoxSH3A-B in complex with p22phox149-168 was obtained from Ogura et al., and the tentative assignment
9
of Trp indole signals of the free p47phoxSH3A-B was derived from titration experiments with p22phox149-168
(Supporting Information Figure S7). Subsequently, titrations with all eight fragment hits were performed
(
p47phoxSH3A-B:fragment ratio 1:0, 1:5, 1:10 and 1:20) and the absolute CSP measured for the indole NH
1
0
ACS Paragon Plus Environment

Page 11 of 64
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
signals during titrations with each of the fragments were characterized (Figure 4D, Supporting Information
Figure S8 and Table S3). All CSPs of the indole NHs of p47phoxSH3A-B upon titration of p22phox149-168 reflect
binding in slow exchange. Fragment hits 1 and 2 were the only fragments showing CSPs for all five indole
NHs of the Trps, including strong CSPs with slow to intermediate exchange binding kinetics for Trp residues
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
93
194
Trp , Trp and Trp264 situated both in the SH3A and SH3B region of p47phoxSH3A-B
204
263
Trp , Trp
(Supporting Information Table S3 and Figure S8A-B). Fragment hits 4, 5 and 7 showed only weak CSP for
1
93
one Trp indole NH (Trp ) located in the SH3A domain, thus indicative of little influence on the magnetic
environment of the indole NHs of p47phoxSH3A-B upon binding to these fragments and binding to only the
SH3A domain (Supporting Information Figure S8G-I, M). Fragment hits 3, 6 and 8 showed CSPs of three or
four indole NHs, mostly characterized as weak.
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Validation of fragment hits by NMR, SAXS and SPR binding stoichiometry. A) The solution NMR
structure of p47phoxSH3A-B (green) in complex with p22phox149-168 (orange) (PDB ID: 1WLP) where the five
Trp indole NH groups of p47phoxSH3A-B situated in the binding pocket are highlighted. B) H- N HSQC
1
15
spectra of p47phoxSH3A-B free (blue) and in the presence of 4-fold molar excess of p22phox149-168 (red). C)
Expanded region with indole NH signals of the five tryptophan residues in p47phoxSH3A-B free (blue) and
bound to p22phox149-168 (red). D) The same expanded tryptophan indole NH region of p47phoxSH3A-B free
1
2
ACS Paragon Plus Environment

Page 13 of 64
1
Journal of Medicinal Chemistry
(blue) superimposed with p47phoxSH3A-B in the presence of fragment hit 1 (1:20) (green), both samples
supplemented with 4% DMSO. E) SAXS data for p47phoxSH3A-B (black) and p47phox in the presence of
p22phox149-168 (grey) plotted as concentration-normalized scattering intensity I(q)/c in arbitrary units vs. q
and F) the corresponding pair-distance distribution function P(r). G) SAXS data for p47phoxSH3A-B (black) and
p47phoxSH3A-B in the presence of fragments plotted as concentration-normalized scattering intensity I(q)/c
in arbitrary units vs. q and H) the corresponding pair-distance distribution function P(r). I) The 8 final
fragment hits were injected by a one-step gradient injection over immobilized p47phoxSH3A-B (2745 RU) to
distinguish the differences in MW normalized response maximum binding level by SPR. J) SPR inhibition
assay where p22phox149-168 was injected constantly at 10 µM while injecting a one-step gradient of 1 mM
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
fragment over immobilized p47phoxSH3A-B
.
Solution structure of p47phoxSH3A-B upon fragment binding investigated by SAXS.
9
10
p22phox-peptides have been shown by Ogura et al. and Yuzawa et al. to bind a compact conformation of
p47phoxSH3A-B. To explore if our eight fragment hits bound in a similar manner, we studied the overall
structure and conformation of p47phoxSH3A-B and the auto-inhibited form p47phoxAIR by SAXS studies. The
pair-distance distribution functions of the SAXS data of p47phoxSH3A-B when binding to p22phox149-168 and
p22phox151-162 are narrower (Figure 4E-F and Supporting Information Figure S9A-B) and the Kratky plots
more bell-shaped (Supporting Information Figure S9C) than for the free p47phoxSH3A-B, thus indicating a
more compact conformation of p47phoxSH3A-B when binding to p22phox peptides and supporting the
9
AIR
conclusions of Ogura et al. A similar compact conformation was observed for p47phox compared to
p47phoxSH3A-B (Supporting Information Figure S9D-F and Table S2), as shown by Yuzawa et al. Possible
alterations of p47phoxAIR upon addition of p22phox-derived peptides were not previously investigated;
however, our data did not indicate any differences between p47phoxAIR with and without p22phox149-168
10
(Supporting Information Figure S9D-F and Table S2). In contrast to the p22phox peptides, no significant
conformational changes in p47phoxSH3A-B were observed with any of the eight fragments at either 2 or 4
mM (Figure 4G-H). One fragment with high solubility (fragment hit 8), was tested also at a 10 mM
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
concentration, showing no difference between any of the concentrations (data not shown). These SAXS
data indicate that the fragments bind to an extended conformation of p47phoxSH3A-B
.
SPR binding level as a measure to elucidate binding stoichiometry.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The NMR experiments of p47phoxSH3A-B with the fragment hits 1 and 2 show an influence of fragments
binding on chemical shifts of Trp residues in both the SH3A and SH3B domains of p47phox. The SAXS data
show that all fragments bind to an extended conformation of p47phoxSH3A-B, and not a compact
conformation as induced upon p22phox binding. This means that the SH3A and SH3B domains of
p47phoxSH3A-B are not close enough in space to form a joint extended binding pocket accommodating
fragment hits 1 and 2, which would explain the chemical shift changes of the Trp residues in both SH3
domains of p47phoxSH3A-B. An alternative explanation could be that two binding pockets for fragments exist
in p47phoxSH3A-B, one in the SH3A domain and one in the SH3B domain. To investigate this, the eight
fragments were injected in a one-step concentration gradient31, 32 at 1 mM over immobilized p47phoxSH3A-B
allowing a direct comparison of the molecular weight (MW) normalized response levels of the eight
fragments (Figure 4I). The response levels of 1 and 2 were approximately 2-fold higher compared to 3-8,
clearly indicating that 1 and 2 have two binding-sites at p47phoxSH3A-B, while 3-8 likely have one, in support
of the NMR results.
Competition between p22phox and fragments for binding to p47phoxSH3A-B
.
The ability of 1-8 to compete with p22phox for binding to p47phoxSH3A-B was investigated in SPR
competition assays. First, this was done in a two-component injection experiment where p22phox149-168 (10
µM) was injected over immobilized p47phoxSH3A-B followed by a continuous gradient injection of fragments
(
up to 1 mM), while the p22phox149-168 concentration was kept constant (Figure 4J). These experiments
were performed directly after the injections shown in Figure 4I, whereby the normalized binding levels in
Figure 4I and Figure 4J can be directly compared. The response levels for all eight fragments binding to
p47phoxSH3A-B was significantly reduced when they were injected with p22phox149-168 (Figure 4J), which
shows that when p47phoxSH3A-B is already binding p22phox149-168, the ability of the fragments to bind
1
4
ACS Paragon Plus Environment

Page 15 of 64
1
Journal of Medicinal Chemistry
p47phoxSH3A-B is reduced. The second competition assay was performed oppositely by injecting the
fragments constantly at 1 mM over immobilized p47phoxSH3A-B and subsequently, whilst continuously
injecting these, adding p22phox149-168 (10 µM) by one-step gradient injection (Supporting Information
Figure S10). Comparing the experimental response levels with the expected theoretical response, based on
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
calculated response if both fragment and p22phox149-168 are able to bind simultaneously to p47phoxSH3A-B
,
revealed that only fragments 1 and 2 were able to reduce the total binding level by more than 5% and thus
compete with p22phox149-168 for binding to p47phoxSH3A-B (Supporting Information Table S4). The results
from the two SPR competition assays illustrate that the ability for competition is influenced by the order of
injection of the competing components. When injecting p22phox149-168 before the fragments were injected,
the p47phoxSH3A-B exists in a more compact peptide-bound conformation and in this case all fragment hits
poorly bind p47phoxSH3A-B. In the experiments where the fragments were continuously injected before
p22phox149-168, we presume that p47phoxSH3A-B exist in a more elongated conformation, and in this case
fragments 1 and 2 were able to reduce p22phox149-168 binding.
SAR studies of fragments 1 and 2.
Fragment hits 1 and 2, which share the quinoline scaffold, were selected for SAR studies (Figure 5A). These
fragments were the only hits originating as common FP and TSA screen hits that were validated in all SPR
assays. Additionally, 1 and 2 showed chemical shift changes of all five Trp indole NHs of p47phoxSH3A-B
.
These NMR data together with the SAXS and SPR data, suggested binding to two sites at the extended
conformation of p47phoxSH3A-B, one in the SH3A and one in the SH3B domain. Twenty-one analogues (9-29)
of fragment 2 and 41 analogues (30-70) of fragment 1 were purchased or synthesized (Scheme 1) and
tested by FP and SPR (Tables 2-3; Figure 5 shows representative data for analogs with reduced, similar or
improved activity). First, the piperazine ring of 2 (R ) was substituted with N-substituted piperazines (9-12),
2
N-heterocyclic rings (13-15), non-cyclic alkyl amines (16, 17), or amino groups (18), while R and R were
1
3
maintained as methyl and hydrogen, respectively. The piperidine and homopiperazine rings turned out to
be favorable in R leading to 1.5–2-fold improvement in affinity of (13) and (15) relative to 2, while most of
2
the N-substituted piperazines (9-11), morpholine (14), the non-cyclic alkyl amine in (16), and amine (18) led
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
to reduced affinities compared to 2. However, the non-cyclic alkyl amine as in 17, containing both a
secondary and a tertiary amine like a piperazine, and N-benzylated piperazine (12) was equipotent to
fragment 2. Next, we found that replacing the methyl group (R ) of 2 with H gave compound 19 with
1
similar affinity as 2. We then used this scaffold to explore substitutions of R including halogens (20, 21),
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
electron-withdrawing trifluoromethyl groups (22, 23) and electron-donating methoxy groups (24, 25) on
the benzene ring of the quinoline. The 6-Br substituted analog (20) stood out from this series and increased
the binding affinity in FP 3-fold compared to 2 or 19. The remaining compounds gave affinities either
similar to 19 (21, 22, 24, 25) or lower (23). Re-scaffolding of the quinoline ring and the position of
piperazine were then investigated. We found that piperazine at position 4 (2, 19) is favored over position 2
(26, 27), as the latter resulted in approximately two-fold decrease in affinity. When the quinoline ring was
replaced with a pyridine scaffold a two-fold increase in affinity was observed by FP for compound 28
relative to (19), while the naphthalene-analog (29) gave a 2.5-fold reduction in affinity (Table 2).
1
6
ACS Paragon Plus Environment

Page 17 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. SAR analysis. A) Fragment hits 1 and 2 share the quinoline scaffold. B) Fragment hit 1 and
analogues 44, 48 and 54 compete with Cy5-p22phox149-162 for binding to p47phoxSH3A-B in the FP assay. C-F)
1
, 44, 48 and 54 tested by SPR for binding to immobilized p47phoxSH3A-B (5873 RU). 44 did not bind to
p47phoxSH3A-B, while 48 binds to p47phoxSH3A-B with a slightly reduced affinity compared to 1. 54 binds to
p47phoxSH3A-B with increased affinity compared to 1. G) The merge analog 64 and parental fragment hits 1
and 2 compete with Cy5-p22phox149-162 for binding to p47phoxSH3A-B in the FP assay. H) SPR sensorgrams of
6
4 injected over immobilized p47phoxSH3A-B. I) Dimeric compound 71 and the p22phox151-161 peptide
compete with Cy5-p22phox149-162 for binding to p47phoxSH3A-B in the FP assay. J) SPR sensorgrams of dimeric
compound 71 injected over immobilized p47phoxSH3A-B
.
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 1. Synthetic procedures for analogs of fragment hit 1 and 2.
H
N
A
B
D
R1
Cl
R3
R1
NH2
N
a or b or c
a
d
R2
R2
N
R2
N
R1
N
R1
N
Cl
7
7
29
74a-74h
9-17, 19-25
75a-75b
26-27, 39-42
7
4a: R1 = CH3, R2 = H
74b: R1 = H, R2 = H
4c: R1 = H, R2 = 6-Br
74d: R1 = H, R2 = 7-Cl
4e: R1 = H, R2 = 7-CF3
74f: R1 = H, R2 = 8-CF3
4g: R1 = H, R2 = 6-OCH3
9
1
1
1
1
1
1
1
1
: R1 = CH3, R2 = H, R3 = N-methylpiperazine
0: R1 = CH3, R2 = H, R3 = N-acetylpiperazine
1: R1 = CH3, R2 = H, R3 = N-Boc-piperazine
2: R1 = CH3, R2 = H, R3 = N-benzylpiperazine
3: R1 = CH3, R2 = H, R3 = piperidine
75a
:
R1 = H
R1 = CH3,
26: R1= H, R2= Piperazine
27: R1=CH3, R2= Piperazine
Cl
N
HN
N
NH2
N
75b
:
7
E
a
e
3
4
4
4
9: R1= H, R2= -NHCH3
0: R1= H, R2= -N(CH3)2
1: R1= H, R2= -NH(CH2)2OCH3
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
4: R1 = CH3, R2 = H, R3 = morpholine
2: R1= H, R2= -NH(CH2)2O(CH2)2OH
7
5: R1 = CH3, R2 = H, R3 = homopiperazine
6: R1 = CH3, R2 = H, R3 = -NH2(CH2)2OCH3
7: R1 = CH3, R2 = H, R3 = -NH2(CH2)2N(CH3)2
74b
78
35
74h: R1 = H, R2 = 6,7-OCH3
19: R1 = H, R2 = H, R3= piperazine
0: R1 = H, R2 = 6-Br, R3 = piperazine
21: R1 = H, R2 = 7-Cl, R3 = piperazine
2: R1 = H, R2 = 7-CF3, R3 = piperazine
23: R1 = H, R2 = 8-CF3, R3 = piperazine
4: R1 = H, R2 = 6-OCH3, R3 = piperazine
C
Cl
N
F
2
a
N
N
NH
f
g-h
i
2
N
N
O
N
NH2
N
NH
76
28
2
O
25: R1 = H, R2 = 6,7-OCH3, R3 = piperazine
7
9
80
81
43
G
Cl
O
Cl
Cl
j
HO
k
Br
l
f
g-h
O
O
O
O
N
N
N
N
N
N
NH2
O
82
83
84
85
86
49
H
I
m
f
g-h
Cl
R
R
Cl
Cl
X
R
N
N
N
O
N
NH2
Br
n
f
g-h
N
N
87a-87c
88a-88g
89a-89g
50-52, 54-56, 58
N
O
N
NH2
8
7a: X = 6-OH
88a: R = 6-O-(p-Cl-Bn)
88b: R = 6-OBn
88c: R = 6-OCH2-cyclohexyl
89a: R = 6-O-(p-Cl-Bn)
89b: R = 6-OBn
89c: R = 6-OCH2-cyclohexyl
89d: R = 6-OCH3
50: R = 6-O-(p-Cl-Bn)
51: R = 6-OBn
52: R = 6-OCH2-cyclohexyl
54: R = 6-OCH3
87b: X = 7-OH
90
91
92
53
87c: X = 8-OH
88d: R = 6-OCH3
88e: R = 6-(CH2)2OCH2CH3
89e: R = 6-(CH2)2OCH2CH3
55: R = 6-(CH2)2OCH2CH3
88f: R = 7-O(CH2)2OCH2CH3 89f: R = 7-O(CH2)2OCH2CH3 56: R = 7-O(CH2)2OCH2CH3
88g: R = 8-O(CH2)2OCH2CH3 89g: R = 8-O(CH2)2OCH2CH3 58: R = 8-O(CH2)2OCH2CH3
63: R = -NH-(p-OCH3-Bn)
65: R = N-methylpiperazine
66: R = N-acetylpiperazine
67: R = piperidine
Cl
R
N
K
Cl
N
Cl
f
g-h
b
J
R
R
R
N
O
NH2 68: R = morpholine
OH
N
O
O
N
N
NH2
69: R = -NH(CH2)2OCH3
O
N
m
f
g-h
70: R = -NH(CH2)2N(CH3)2
7
4b
96
60
63, 65-70
N
O
NH2
g
H
N
H
N
93
94a-94b
94a: R = -(CH2)2OCH2CH3 95a:
95a-95b
61-62
Cl
N
N
N
R = -(CH2)2OCH2CH3
R = -O-(p-Cl-Bn)
61: R = -(CH2)2OCH2CH3
62: R = -O-(p-Cl-Bn)
N
9
4b: R = -O-(p-Cl-Bn)
95b:
a
h
N
H
N
N
NH2
H
97
98
64
Reagents and conditions: (a) 2-propanol, amine, reflux, 12–24 h; (b) NMP, amine, microwave heating, 200
ο
ο
C, 15 min; (c) K CO , homopiperazine, DMF, 120 C, 24 h; (d) bis(2-chloroethyl)amine hydrochloride, K CO
2
3
2
3,
ο
diglyme, reflux, 48 h; (e) TFA, rt, overnight; (f) m-CPBA, DCM, 0 C to rt, overnight; (g) tert-butylamine, p-
ο
toluenesulfonic anhydride, trifluorotoluene, DCM, 0 C to rt, 30 min; (h) TFA, reflux, 3 h; (i) acetyl chloride,
ο
ο
triethyl amine, DMAP, DCM, 0 C, 2h, then excess methylamine, rt, 16 h; (j) LiAlH , THF, -20 C, 2 h; (k) PBr ,
4
3
ο
ο
DCM, 0 C to reflux, 2 h; (l) K CO , 4-chlorophenol, DMF, 80 C, 16 h; (m) K CO , organobromide, DMF, 80
2
3
2
3
ο
C, 16 h; (n) (4-chlorophenyl)boronic acid, K CO , Pd(PPh ) , dioxane/H O, reflux, 2 h; 89e was commercially
2
3
3 4
2
available.
1
8
ACS Paragon Plus Environment

Page 19 of 64
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. SAR analysis of fragment hit 2 analogs.
R2
R3
N
R1
(µM)a
(µM)b
SPR K
D
Compds
2
R
1
R
2
R
3
FP K
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CH
CH
CH
3
3
N
N
N
NH
N
H
H
H
606 ± 25
> 2000
412  130
9
W
10
3
3
N
815 ± 96
W
O
O
O
1
1
1
2
CH
H
> 2000
NB
N
N
CH
CH
CH
3
3
3
N
N
N
N Bn
H
H
H
1000 ± 480
403 ± 20
W
13
470
597
1
1
4
5
O
1600 ± 110
NH
CH
CH
3
3
N
H
H
271 ± 32
296
997
16
NH
NH
O
1120 ± 23
1
1
7
8
CH
CH
H
3
3
N
H
H
H
722 ± 39
1070 ± 88
697 ± 77
332
521
487
NH
2
19
N
N
N
N
N
N
N
NH
NH
NH
NH
NH
NH
NH
2
0
H
H
H
H
H
H
6-Br
7-Cl
7-CF
8-CF
231 ± 6.6
583 ± 24
NT
W
21
2
2
3
3
894 ± 35
W
23
1540 ± 230
466 ± 37
W
2
4
6-OCH
3
345
318
420
W
25
6,7-OCH
3
598 ± 62
2
6
N
NH
NH
H
H
H
1410 ± 84
1070 ± 130
302 ± 20
27
N
CH
3
2
8
N
N
NH
NB
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
H
N
2
9
N
1520 ± 80
366
aK values are reported as the mean ± SEM (n ≥ 3); bK
values are reported based on n = 1 or n = 2-3 (mean ± SEM).
i D
NB = No binding (MW norm response max<20% compared to parental fragment hit 2), W = weaker binding (MW norm
response max>20%, <50% compared to parental fragment hit 2), NT = Not tested.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 3. SAR analysis of fragment hit 1 analogs.
R2
R3
N
R1
SPR K
D
b
Compds
1
R
1
R
2
R
3
FP K
i
(µM)a
(
µM)
NH
2
H
H
428 ± 23
> 2000
488  130
NB
3
0
H
H
H
31
CH
3
H
H
> 2000
W
3
3
2
3
OH
H
H
> 2000
NB
-CH
H
2
NH
2
H
H
> 2000
NB
34
H
5-NH
H
2
> 2000
W
3
5
H
NH
2
1520 ± 160
> 2000
765
NB
36
H
H
6-NH
8-NH
2
2
3
7
H
H
> 2000
NB
NH2
38
39
> 2000
W
N
NH
N
H
H
H
H
1370 ± 220
1660 ± 140
W
W
4
4
0
1
O
O
H
N
H
H
H
H
H
H
> 2000
> 2000
> 2000
> 2000
NB
NB
NB
NB
H
N
4
2
OH
O
43
NH
4
4
N
NH2
2
0
ACS Paragon Plus Environment

Page 21 of 64
Journal of Medicinal Chemistry
1
2
N
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
45
> 2000
> 2000
> 2000
NB
NB
NB
N
NH2
4
4
6
7
NH2
S
N
NH2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
4
8
594 ± 8.7
687
NH2
4
5
5
5
9
0
1
2
NH
NH
NH
NH
2
2
2
2
H
H
H
H
6-CH O
Cl
Cl
> 2000
> 2000
> 2000
> 2000
NB
NB
W
2
6-OCH2
6-OCH2
6-OCH2
W
5
5
5
5
5
3
4
5
6
7
NH
NH
NH
NH
NH
NH
NH
NH
2
2
2
2
2
2
2
2
H
6-(p-Cl-Ph)
> 2000
241 ± 16
373 ± 14
254 ± 8.5
1380 ± 190
1250 ± 120
> 2000
NB
341
H
6-OCH
3
H
6-O(CH
7-O(CH
8-OH
8-O(CH
H
2
)
2
OCH
OCH
2
CH
CH
3
3
562 ± 56
599 ± 68
*
H
)
2 2
2
H
58
H
)
2 2
OCH
2
CH
3
987
5
9
OH
Cl
NB/W
W
60
H
> 2000
O
O
6
1
NH
2
H
512 ± 22
153
O
Cl
O
6
2
NH
2
H
> 2000
W
H
N
6
3
NH
2
H
243 ± 2.4
NT
6
6
6
4
5
6
NH
NH
NH
NH
NH
2
N
N
N
NH
N
H
H
H
H
H
196 ± 15
597 ± 42
475 ± 19
186 ± 11
435 ± 13
78 ± 39
643 ± 170
820
2
N
2
O
67
2
2
N
N
109
6
8
O
140
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
7
9
0
NH
2
2
NH
O
N
H
H
658 ± 53
592 ± 40
97
NH
NH
O
574 ± 4
O
O
N
O
O
O
N
7
1
20.0 ± 0.95
50 ± 3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH2
H N
2
O
O
N
N
N
N
7
2
30.0 ± 3.2
32.7 ± 2.9
177 ± 16
N
NH2
O
H N
N
2
O
O
n = 3
73
*
H N
2
N
N
NH2
values are reported based on n = 1 or n = 2-3 (mean ± SEM). NB
No binding (MW norm response max<20% compared to parental fragment hit 1), W = weaker binding (MW norm
aK values are reported as the mean ± SEM (n ≥ 3); bK
=
response max>20%, <50% compared to parental fragment hit 1, NT = Not tested, *ambiguous SPR binding curve.
i
D
For 1, close analogues revealed that the free amino group (R ) is crucial for activity, as replacements (30-
1
3
3), position change (34-38), or alkylation/acetylation (39-43) resulted in almost complete loss of affinity.
More dramatic changes of the quinoline scaffold to pyridine, pyrimidine, naphthalene, or benzo[d]thiazole
also resulted in inactive compounds (44-47), while the isoquinoline compound (48) showed only a 1.5-fold
reduction in affinity compared to 1. Based on this, further modifications were carried out on the conserved
2
-aminoquinoline binding core. Noticeably, 6-substituted-2-aminoquinolines have been reported in
literature as small-molecule ligands for Tec SH3 protein domains with 6-[(4-
chlorophenoxy)methyl]quinolin-2-amine as the most potent compound showing a K of 9 µM by NMR.
3
3-35
D
Inspired by this, we investigated various substitutions on the 6-position of the quinoline ring including the
mentioned literature compound (49). However, in contrast to the published activity against Tec SH3, 49
was not active by either FP or SPR measuring inhibition of the p47phox/ p22phox interaction and binding to
p47phox, respectively, and it also had poor solubility properties. Small modifications based on 49 were
conducted in an attempt to improve affinity, but the resulting compounds (50-53) were also weak binders
with solubility issues like 49. However, substitutions with methoxy (54) or linear alkoxy groups (55-56) on
2
2
ACS Paragon Plus Environment

Page 23 of 64
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the 6- and 7-position of 1 resulted in same or two-fold improved affinities compared to 1, while
substitutions on the 8-position with hydroxy (57) or an alkoxy-group (58) were not well-tolerated resulting
in decreased affinity. Instead, we studied the SAR around the 4-position of 1 (R ), where it was observed
2
that small substitutions such as OH (59) and Cl (60) were not favored, while the linear alkoxy group as in
(61) restored the affinity. Installing the same substitution group as found in the ‘literature compound’ (49),
but here in the 4-position, gave the inactive compound 62, while the substitution of 4-methoxybenzylamine
gave compound 63 showing a two-fold improvement in affinity relative to 1. Considering the importance of
the piperazine ring in R of 2 (2 vs 31) and the amino group in R of 1 (1 vs 30) we expected the merge of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1
the two hits 1 and 2 to show improved affinity, and indeed compound 64 demonstrated 2–3-fold lower Ki
values by FP relative to parent fragment hits and also showed enhanced binding by SPR (K = 78 µM)
D
(Figure 5G-H, Table 3). We thus elaborated on this merged scaffold with compounds 65-70 by modifying or
substituting the piperazine ring similar to what we did for the analogs of 2 (Table 3). Interestingly, some of
the same trends were seen, as methylation or acetylation of the piperazine (compound 65 and 66,
respectively) reduced the affinity, as did ring-opening of the piperazine to compound 69 and 70, while the
piperidine compound 67 gave a two-fold increased affinity. In this case, the morpholine substitution
resulted in only a two-fold reduction in affinity (68 vs 1) (Table 3).
Overall, several analogs showed improved affinities relative to parent fragment hit 1 and 2, however, to
a rather modest extent with 2-3-fold decrease in K values for compound 13, 15, 20, 28, 54, 56, 63, 64 and
i
6
7. Thus, it seemed that to obtain more dramatic improvements in inhibitory activity and binding affinity,
more unconventional means were required. Hence, considering that our SPR and NMR data (Figure 4 and
Supporting Information Figure S8) suggest that 1 and 2 bind to both SH3 domains of p47phox, we
envisioned that connecting these fragments by a flexible linker able to span both SH3 domains affording a
homodimeric compound would lead to more significant affinity improvements, as has been reported for
other protein targets.36, 37 Based on our SAR study, the 4-, 6-, 7-positions could provide suitable attachment
points for linkers. Thus, the three dimeric compounds 71-73 were designed and synthesized (Scheme 2).
Encouragingly, the three dimers gave a 13–30-fold improved affinity by FP compared to the monomeric
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
parent fragment hits 1 and 2 (Table 3). Among the three dimeric compounds, 71 was the most potent
151-162
dimer with a K of 20 µM in FP, which is similar to the p22phox
peptide (K = 17 μM at 2% DMSO), and
i
i
with confirmed binding in SPR (K = 50 µM) (Figure 5I-J). As a control experiment for potential aggregation
D
artifacts, 71-73 were tested by FP in the presence of 0.01% Triton-X and showed same affinities as without
the presence of Triton-X within the experimental error (Supporting Information Figure S11).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Synthesis of dimeric p47phox/p22phox inhibitors compound 71, 72, and 73.
O
O
A
O
O
O
O
O
N
O
O
O
O
O
OH
O
N
O
O
N
a
b
c-d
N
NH2
H2N
N
O
N
O
N
71
9
3
99
100
B
O
O
O
O
O
O
H
N
N
N
N
N
N
N
N
e
d
N
N
N
N
N
N
H
N
H
N
N
O
NH2
H2N
N
H
98
101
72
C
O
O
HO
f
O
O
O
c-d
O
b
O
O
n = 3
n = 3
N
O
n = 3
N
O
H2N
N
N
NH2
N
N
N
73
87a
102
103
ο
Reagents and conditions: (a) Cs CO , 1,11-dibromo-3,6,9-trioxaundecane, DMF, 80 C, 2 h; (b) m-CPBA,
2
3
ο
ο
DCM, 0 C to rt, overnight; (c) tert-butylamine, p-toluenesulfonic anhydride, trifluorotoluene, DCM, 0 C to
rt, 30 min; (d) TFA, reflux, 3 h; (e) tetraethylene glycol, K CO , rt, overnight; (f) tetraethylene glycol, PPh ,
2
3
3
DIAD, THF, rt, 72 h.
Binding model.
The competition assays show that fragments 1 and 2 could reduce binding of p22phox to p47phox, and
SPR, NMR, and SAXS data suggest that 1 and 2 bind to both SH3 domains of the extended conformation of
p47phox. We propose that inhibition could be either due to the fragments sterically preventing p22phox149-
1
68
binding or because the fragments stabilize the extended conformation of p47phoxSH3A-B and prevent the
formation of the compact conformation required for creating the binding pocket to p22phox149-168 (Figure
6
A). Based on this, we designed the dimeric compounds 71–73, and the improved affinity seen for these
supports the proposed binding mode of the original fragments. Likewise, these data support that the
2
4
ACS Paragon Plus Environment

Page 25 of 64
1
Journal of Medicinal Chemistry
dimeric compounds occupy both SH3 domains in p47phoxSH3A-B in a binding mode that is likely facilitated by
the interdomain flexibility of p47phoxSH3A-B and the conformational flexibility of the PEG-linker of the
compounds (Figure 6B). Interestingly, NMR binding studies of 2-aminoquinoline with the mouse Tec Kinase
SH3 single domain indicated that the two residues Trp215 and Asp196 in the Tec kinase SH3 sequence are
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
important for binding. Similar residues are found in p47phoxSH3A-B, as Trp215 is conserved for SH3A (Trp193)
3
4
263
196
174
and SH3B (Trp ), while Asp is conservatively exchanged to a glutamic acid in p47phox SH3A (Glu ) and
2
54
SH3B (Glu ) (Supporting Information Figure S12). Combined with our data, this homology further
supports that fragments (1 and 2) and dimeric compounds (e.g. 71) bind to p47phoxSH3A-B at sites composed
1
74
193
254
263
of key amino acids at both SH3A (Glu , Trp ) and SH3B (Glu , Trp ) (Figure 6).
A
B
Figure 6. Model of fragments and dimeric compounds binding to p47phox. A) When blocking the tandem
SH3 domains of p47phoxSH3A-B with the p22phox-derived peptide p22phox149-168 the fragments’ ability to
bind p47phoxSH3A-B was significantly reduced. Fragment hits 1 and 2 were found to bind one site in SH3A
and another site in SH3B of the extended conformation of p47phoxSH3A-B. Binding of 1 and 2 to p47phoxSH3A-
B
reduced p22phox149-168 binding either due to steric hindrance or by stabilizing the extended conformation
of p47phoxSH3A-B. B) Proposed binding mode of dimeric compounds to p47phoxSH3A-B
.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
CONCLUSION
Reduction of NADPH oxidase 2 activity by inhibiting the p47phox-p22phox interaction is a potential strategy
for therapeutic intervention of oxidative stress. To date, ebselen is the only small-molecule shown able to
bind p47phox and prevent the interaction with p22phox.13
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Here, we demonstrate that ebselen covalently reacts with Cys196 of p47phoxSH3A-B, leading to
destabilization and aggregation of the protein; a mechanism of action that is not promising for further drug
development, nor warrant its use as a chemical probe. Instead, we identified eight fragment hits (1–8)–by
screening 2,500 fragments using FP and TSA followed by hit validation by SPR–which bind p47phoxSH3A-B and
serve as scaffolds for developing reversible and druglike small-molecule inhibitors. The quinolone fragment
hits 1 and 2 were the only common hits across all assays and were able to reduce the binding of p22phox to
p47phoxSH3A-B. Binding as well as structural NMR data suggested that they interact with two sites on
p47phoxSH3A-B, presumably at the SH3A and SH3B domain, and interestingly, SAXS studies suggested that 1
and 2 bind p47phoxSH3A-B when this is in an extended conformation instead of the compact conformation
seen when p22phox peptide binds.
Sixty-five analogues of 1 and 2 were designed and synthesized. Several smaller modifications gave
affinity improvements of 2–3-fold compared to parent fragment hits. However, the biggest improvement in
affinity was obtained by connecting fragments 1 and 2 by a PEG-based linker, affording the three
homodimeric compounds 71-73 with 13–30-fold higher affinities than 1 and 2, and the best dimer (71)
being equipotent to the p22phox151-162 peptide (K = 20 μM). This dimeric design was incited by our SPR,
i
SAXS, and NMR data suggesting that 1 and 2 bind to both SH3 domains of p47phox, and the improved
affinity seen for the dimeric compounds support the proposed binding mode of the fragments.
By fragment-based drug discovery, we have developed the first non-peptide reversible inhibitors of the
p47phox-p22phox protein-protein interaction. The best compounds are dimeric with affinities similar to the
shorter p22phox-derived peptide p22phox151-162. However, further optimization is needed to reach affinities
relevant for more advanced biological studies, such as cell-free or cell-based ROS detection assays and cell-
based oxygen consumption measurements.14, 38 Our SAR study and proposed binding model based on
biophysical and structural data provide encouraging results and guidance for future directions. Hopefully
2
6
ACS Paragon Plus Environment

Page 27 of 64
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
these data can facilitate development of chemical probes suitable for studying the therapeutic relevance of
NADPH oxidase 2 inhibition.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EXPERIMENTAL SECTION
Chemistry. General Procedures. All commercial available solvents and reagents were purchased and
used without further purification. Room temperature refers to 20–22°C. Normal-phase chromatography
was executed on TELEDYNE ISCO apparatus (CombiFlash Rf), using silica gel as the stationary phase and
mixtures of ethyl acetate and heptane or dichloromethane (DCM) and methanol (MeOH) as eluents.
Preparative reverse-phase chromatography was performed on Agilent 1200 series HPLC system using a C18
reverse column (ZORBAX 300SB-C18 PrepHT, 21.2 × 250 mm, 7 µm) with a linear gradient of the binary
solvent system of H O⁄CH CN⁄TFA (A: 95⁄5⁄0.1, and B: 5⁄95⁄0.1) with a flow rate of 20 mL⁄min. Microwave
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3
heating was performed on Biotage Initiator 2.5. Nuclear magnetic resonance (NMR) spectra were recorded
on 600 MHz Bruker Avance III HD instrument equipped with a cryogenically cooled 5 mm dual probe or 400
MHz Bruker Avance II instrument equipped with a 5 mm broadband probe. Deuterated dimethyl sulfoxide
1
13
(
DMSO-d ) or deuterated chloroform (CDCl ) were used as solvents. Chemical shifts for H and C spectra
6
3
were reported in parts per million (ppm) and coupling constants were reported in Hertz. LC-MS analyses
were carried out using an Agilent 6410 Triple Quadrupole Mass Spectrometer instrument with electron
spray ionization (ESI) coupled to an Agilent 1200 HPLC system (ESI-LC/MS) with a C18 reverse phase column
(Zorbax Eclipse XBD-C18, 4.6 mm × 50 mm), autosampler and diode array detector using a linear gradient of
the binary solvent system of H O⁄CH CN⁄formic acid (A: 95⁄5⁄0.1, and B: 5⁄95⁄0.043). Also, evaporative
2
3
light scattering (ELS) traces were obtained with a Sedere Sedex 85 Light Scattering Detector. The LC-MS
method for all the target compounds is as follows: ESI, 0% B to 100% B, run time = 5 min, flow rate = 1
mL/min. Compounds 18, 30-34, 36-38, 44-48, 57 and 59 were purchased and showed ≥ 95% purity by NMR
(1
H, 13C) and LC-MS (UV). All other final compounds were synthesized and showed ≥ 95% purity, as
1
13
1
1
1
13
determined by NMR ( H, C, H- H COSY, H- C HSQC) and LC-MS (UV).
Peptides. TAMRA labeled peptide p22phox151-162 (5(6)TAMRA-PPTNPPPRPPAE-NH2) and unlabeled
peptides - p22phox151-162 (Ac-PPTNPPPRPPAE-NH2), p22phox149-168 (Ac-KQPPSNPPPRPPAEARKKPS-NH2), and
p22phox149-162 (Ac-CKQPPSNPPPRPPAE-NH2) - were purchased from Biomatik. The Cy5 labeled peptide
2
8
ACS Paragon Plus Environment

Page 29 of 64
1
Journal of Medicinal Chemistry
p22phox149-162 with sequence Ac-C(Cy5)KQPPSNPPPRPPAE-NH2 was generated by mixing unlabeled peptide
Ac-CKQPPSNPPPRPPAE-NH2 (2.24 mg) dissolved in 1.7 mL buffer (50 mM Hepes, 150 mM NaCl, pH 7.4)
with sulfo-Cy5-maleimide (1.14 mg; Lumiprobe) dissolved in 40 μL DMSO followed by incubation for 2
hours, HPLC purification, and lyophilization. All peptides were analyzed by LC-MS and showed ≥ 95% purity.
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General Procedure A for the Synthesis of 9-14, 17, 19, 21-23, 25-28, 39-42, 78. A mixture of properly
substituted 4-chloroquinoline or 2-chloroquinoline (1 equiv) and amines (2 equiv) in isopropyl alcohol (3
mL) was refluxed for 12-24 h. The solvent was evaporated in vacuo, and the resulting residue was purified
by silica gel column chromatography (9-14, 19, 21-23, 25-28, 39-42, 78) or reverse-phase (RP)-HPLC (17) to
obtain the desired product.
General Procedure B for the Synthesis of 16, 20, 24, 63, 65-70. A mixture of properly substituted 4-
chloroquinoline or 4-chloroquinoline-2-amine (1 equiv) and amines (2 equiv) in NMP (1 mL) was heated at
2
00 °C for 15 min in a microwave reactor. After cooling to room temperature, the reaction mixture was
diluted with water and extracted three times with ethyl acetate. The combined organic layers were washed
with brine, dried over Na SO , filtered and concentrated. The residue was purified by silica gel column
2
4
chromatography (20, 63, 65-69) or RP-HPLC (16, 24, 70) to afford the desired product.
General Procedure C for the Synthesis of 80, 86, 89a-89g, 92, 95a-95b, 96, 100, 103. To a solution of
ο
properly substituted quinoline (1 equiv) in DCM (15 mL) at 0 C was added mCPBA (1.5 equiv or 3 equiv).
The reaction mixture was allowed to warm to room temperature and stirred overnight. The reaction was
quenched with 1M NaOH until pH = 10. The aqueous layer was extracted with DCM three times and the
combined organic extracts were dried over Na SO , filtered and concentrated under pressure. The crude
2
4
product was used without further purification, unless otherwise stated.
General procedure D for the synthesis of 49-56, 58, 60-62, 71, 73, 81. The properly substituted
quinoline 1-oxide (1 equiv) was diluted in 2:1 trifluorotoluene/DCM and cooled to 0 °C. tert-Butylamine (5
2
9
ACS Paragon Plus Environment